--DCVUsB8wYFFoocrD
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri, Jul 13, 2007 at 03:20:12PM +0200, Matus UHLAR - fantomas wrote:
> On 13.07.07 09:15, Matt Kettler wrote:
> > It should only match a missing header.


"missing header" doesn't mean "the header doesn't exist", btw. It's a subt=
le
difference, please see below.

> bug 5207 should be reopened then...


Nope.

I'll try to keep it short, but here's what's going on:

A header is parsed as:
"[header_name]:[optional_whitespace][header_data]"

The header "exists:header_name" test really converts into:
header_data =3D~ /./

which requires that there be a character in the header_data.

MISSING_SUBJECT is the inverse of the "exists:Subject" rule, made using
a meta rule.

If the Subject header is actually missing, then there is no header_data, and
therefore the "/./" test fails.

If the Subject header exists, but has nothing but whitespace after the
colon ... there's no header_data, and therefore the "/./" test fails.


I'll put this in the bug for future reference.

--=20
Randomly Selected Tagline:
"Those who do not archive the past are condemned to retype it!"
- Garfinkel and Spafford

--DCVUsB8wYFFoocrD
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)

iD8DBQFGl62FamwUIkXWD1cRAja7AJ0QGSQq0dloOBoFklmQGF Gxcef5AQCfdwcQ
KbJyDkxRzUkPfX8G2CuUbjE=
=JWhV
-----END PGP SIGNATURE-----

--DCVUsB8wYFFoocrD--